[prev in list] [next in list] [prev in thread] [next in thread] 

List:       shibboleth-users
Subject:    Re: Error 404 after processing LogoutResponse from IdP on SP
From:       Robert Ayrapetyan <robert.ayrapetyan () gmail ! com>
Date:       2014-12-27 23:22:31
Message-ID: 549F3F37.2080404 () gmail ! com
[Download RAW message or body]

Created issue - https://issues.shibboleth.net/jira/browse/SSPCPP-637.
Thanks!

On 12/27/2014 12:30, Cantor, Scott wrote:
> On 12/26/14, 10:27 PM, "Robert Ayrapetyan" <robert.ayrapetyan@gmail.com>
> wrote:
>
> > It doesn't look easy to submit a bug/issue (all the ways found end up
> > with a "Contact JIRA admin" or selection of IDP-organizations I'm not
> > belonging to).
>
> You can get an account at ProtectNetwork if you don't have an account
> anywhere, which it says right on the discovery page.
>
> > I've checked logs and screenshots I've provided - they are really
> > messed-up (it must be SAMLResponse there, not SAMLRequest, sorry about
> > that). The very first message however looks right (all steps are
> > described correctly).
>
> It clearly says SAMLRequest, which is what I noted. So if nothing else,
> the IdP here is broken, which hardly inspires confidence.
>
> I doubt the SP will notice that specific minor detail of the binding spec
> and probably won't care if it sends a SAML response with the wrong
> parameter name, but the point is, that means the IdP is broken. Which
> means it could be broken in other ways that will turn out to be relevant
> to the problem.
>
> > To simplify things, just check what happens after IDP replies with
> > "LogoutResponse" to IIS-based SP (url /Shibboleth.sso/SLO/Redirect) - you
> > will always see this 404 error.
>
> If you file a bug, then when I do have time, I will. I still don't believe
> the SP itself ever returns a 404, which means whatever this is, it's
> probably environmental.
>
> -- Scott
>

-- 
To unsubscribe from this list send an email to users-unsubscribe@shibboleth.net
[prev in list] [next in list] [prev in thread] [next in thread]